qfq/s1K_v2_rep6

该数据集包含了多个字段，其中包括解决方案(solution)、问题(question)、上下文类型(cot_type)、来源类型(source_type)、元数据(metadata)、上下文(cot)、思维轨迹(thinking_trajectories)和尝试(attempt)。数据集被分割为训练集(train)，共有1000个示例，大小为16716317.738711914字节。数据集提供了一个默认配置，用于指定训练数据的文件路径。

The dataset includes multiple fields such as solution, question, context type (cot_type), source type (source_type), metadata, context (cot), thinking trajectories, and attempt. The dataset is split into a training set (train) with a total of 1000 examples and a size of 16716317.738711914 bytes. A default configuration is provided to specify the file path for the training data.